Yield Curve: Shapes, Interest Rate Expectations, and Economic Signals
A yield curve shows how interest rates differ across debt maturities at a given time. A normal curve generally slopes upward, a flat curve offers little extra yield for longer terms, and an inverted curve places some short-term yields above long-term yields. The shape reflects policy expectations, inflation, term premiums, and demand for bonds—not a guaranteed recession forecast.
What the Yield Curve Measures—and What It Does Not
The Treasury yield curve compares the annualized yields on U.S. government obligations with different maturities. The horizontal axis represents time to maturity, while the vertical axis represents yield. Analysts commonly compare three-month bills, two-year notes, ten-year notes, and thirty-year bonds. They can also examine the entire curve rather than reducing it to one spread.
Importantly, a government yield curve is not the same as a corporate borrowing curve. Treasury instruments carry very little U.S.-dollar default risk relative to ordinary corporate bonds. By contrast, corporate yields contain credit compensation, which can rise independently of Treasury rates. Therefore, a spread between a corporate bond and a Treasury security measures something different from the slope of the Treasury curve.
The U.S. Treasury publishes constant-maturity par yields at standardized points. Its methodology uses indicative market quotations for recently auctioned securities and estimates yields at fixed maturities. As a result, a published ten-year rate is a consistent curve measure; it need not equal the yield of one particular bond held by an investor.
Three shapes, plus an important variation
| Curve shape | Typical yield relationship | Possible interpretation | Important limitation |
|---|---|---|---|
| Normal or upward-sloping | Long-term yields exceed shorter yields | Compensation for duration risk and/or expectations for future rates | A positive slope does not guarantee economic growth |
| Flat | Short and long yields are similar | Changing policy expectations or declining term premiums | A flat reading can occur during several economic conditions |
| Inverted | At least one short maturity yields more than a longer maturity | Expected easing, restrictive current policy, or technical demand | The signal cannot determine recession timing by itself |
| Humped | Intermediate maturities yield more than either end | Rate expectations and risk compensation differ across horizons | One two-point spread misses the shape altogether |
Why Longer Bonds Usually Pay Different Yields
A useful analytical approximation separates a long-term nominal yield into two parts: the market’s expected average path of future short-term rates and a term premium. The premium compensates investors for the risks and characteristics of holding longer-term bonds. Because expectations and premiums both move, analysts should resist interpreting every change in the ten-year yield as a change in expected central-bank policy.
For example, imagine the market expects the average short-term rate over the coming ten years to be 3.2%, while the estimated term premium is 0.8 percentage points. A simplified decomposition gives a ten-year yield of about 4.0%. If the premium rises to 1.3 points and rate expectations remain unchanged, the ten-year yield may increase toward 4.5%. This is an illustrative calculation, not a measured market estimate.
Policy expectations influence the front and middle
Central banks exert their strongest direct influence on overnight interest rates. Markets then price expectations for policy decisions into securities with longer maturities. Therefore, two-year yields often respond to changing expectations about policy over the next several years. However, they also embed risk and market-pricing effects.
Term premium estimates are models, not directly observed facts
Researchers cannot read the term premium off a trading screen. Instead, they estimate it using bond prices, models, and sometimes survey expectations. Federal Reserve researchers illustrated this uncertainty when analyzing the 2023 rise in Treasury yields. For October 19, 2023, the ten-year term premium estimates in their comparison included approximately 0.39 percentage point under one model and 1.37 percentage points under another.
Those values describe the researchers’ specific models and date—not a stable premium investors should apply today. Nevertheless, their disagreement conveys a durable lesson: yield-curve interpretation depends partly on assumptions about expectations and risk compensation. A confident recession narrative based on one decomposition deserves scrutiny.
How to Calculate the Yield-Curve Spread
The simplest way to measure slope is to subtract a shorter-maturity yield from a longer-maturity yield. For the ten-year and two-year comparison:
10-year/2-year spread = 10-year Treasury yield − 2-year Treasury yield.
Suppose a ten-year Treasury yields 3.90% and a two-year Treasury yields 4.25%. The spread equals 3.90% − 4.25% = −0.35 percentage point, or −35 basis points. One basis point equals 0.01 percentage point. Because the spread is negative, this segment is inverted.
Next, suppose the three-month yield is 4.60%. The ten-year/three-month spread equals 3.90% − 4.60% = −0.70 percentage point, or −70 basis points. Both measures are inverted, but their magnitudes differ. These figures are hypothetical and do not represent rates on a particular date.
| Illustrative maturity | Annualized yield | Comparison with ten-year rate |
|---|---|---|
| Three-month | 4.60% | 10y–3m = −70 bp |
| Two-year | 4.25% | 10y–2y = −35 bp |
| Five-year | 4.05% | 10y–5y = −15 bp |
| Ten-year | 3.90% | Reference maturity |
| Thirty-year | 4.20% | 30y–10y = +30 bp |
This invented curve is not uniformly inverted: the thirty-year yield exceeds the ten-year yield. The example shows why one headline about “inversion” cannot fully describe rates across the maturity spectrum.
Use consistent data rather than mixing quotations
First, choose the same currency, issuer, observation date, and yield convention. Second, calculate your spread from rates that use compatible definitions. Finally, note whether you are using daily observations or monthly averages. The Treasury’s par curve quotes should not be casually mixed with a discount-basis bill rate or a different vendor’s zero-coupon estimate.
The Treasury revised its curve methodology on December 6, 2021. Therefore, analysts comparing long samples should document construction changes rather than assume every historical rate was estimated identically.
Why Investors Watch the Ten-Year/Three-Month and Ten-Year/Two-Year Spreads
The ten-year/two-year spread appears frequently in financial commentary because both points are liquid, familiar benchmarks. However, research on recession forecasting often favors the ten-year/three-month spread. The two measures overlap, yet they can turn negative at different times.
In a 2006 Federal Reserve Bank of New York study, Arturo Estrella and Mary Trubin reported that the ten-year/two-year spread tended to invert earlier and more often than their preferred ten-year/three-month measure. Their historical sample also found that, on average, the ten-year/three-month spread was about 21 basis points when the ten-year/two-year spread stood at zero. The precise figure is sample-specific; the general implication is that different maturity pairs cannot be treated as interchangeable.
| Indicator | What it measures | Common use | What may distort the signal |
|---|---|---|---|
| 10y–2y spread | Long Treasury yield versus two-year yield | Headline slope and shifts in medium-term policy expectations | Different risk premiums and earlier/more frequent inversions |
| 10y–3m spread | Long Treasury yield versus very short bill rate | Historically tested recession-risk indicator | False positives, model specifications, and timing uncertainty |
| Near-term forward spread | Implied forward short rate versus a current short rate | Market-implied path of policy over the near term | Forward-rate construction and model interpretation |
| Entire maturity curve | Yields across many dates | Duration pricing, curve trades, and risk assessment | Liquidity differences and market segmentation |
Furthermore, a Federal Reserve Board research paper by Eric Engstrom and Steven Sharpe found that a near-term forward spread outperformed conventional long-term spreads in their statistical models for U.S. recessions and growth. The paper associated the forward measure with expectations for monetary easing over roughly the next year and a half. Its finding does not mean the forward spread will always win in every future sample. Instead, it shows why analysts should examine the rate path priced by markets, not just one popular difference.
Does an Inverted Yield Curve Predict a Recession?
An inverted Treasury yield curve has historically carried useful information about U.S. recession risk. Nevertheless, inversion is a probabilistic warning, not a causal law or a countdown clock. Monetary tightening may push short rates higher, while expectations for weaker growth and future rate cuts hold long rates down. Those conditions can accompany a slowdown, but policy can change and other forces can intervene.
The New York Fed publishes a model that translates the ten-year/three-month spread into an estimated probability of recession twelve months ahead. The bank explicitly says those estimates are not official recession forecasts of the institution or the Federal Reserve System. Therefore, readers should not interpret a model output as a policy committee’s prediction.
What historical research actually found
The New York Fed’s 2006 study offers several unusually practical results. In the study’s 1968–2006 historical window, its preferred monthly spread inverted before each of six recessions. Moreover, each episode featured at least three negative monthly averages during the twelve months before the recession began. By contrast, daily negative spreads appeared on 100 days between January 1968 and December 2005 in months whose overall average did not invert.
Consequently, a few hours of inversion should not carry the same weight as a persistent signal. The study also reported that model-implied recession probabilities exceeded 30% ahead of each recession in its sample, reaching 98% around the 1981–1982 episode. These are historical, model-dependent outcomes from an older sample; they are not universal probabilities for today’s economy.
| Recession episode (historical study) | Negative monthly spreads in prior 12 months | Lowest 10y–3m spread |
|---|---|---|
| 1970 | 10 months | −51 bp |
| 1973–1975 | 6 months | −159 bp |
| 1980 | 12 months | −220 bp |
| 1981–1982 | 10 months | −351 bp |
| 1990–1991 | 3 months | −8 bp |
| 2001 | 7 months | −70 bp |
Historical sample only: figures are from Estrella and Trubin’s 2006 New York Fed study and describe the twelve months preceding each listed recession, not predictions for future episodes.
Other research documents false positives in the late 1960s and 1990s. Historical association cannot identify the exact month of a future downturn.
Persistence matters more than a dramatic intraday headline
Suppose the 10y–3m spread falls to −5 basis points for one trading session and returns above zero the next week. That episode differs from a sustained monthly average of −70 basis points. The first may reflect technical trading or temporary demand, while the second suggests a more persistent repricing of monetary and economic expectations.
Analysts should therefore track daily movements for context but rely on a consistent monthly series when comparing a new inversion against historical recession studies. The exact averaging rule must match the research being quoted. Otherwise, analysts risk overstating the precision of their historical comparison.
Why the Curve Can Be Wrong—or Appear to Be Wrong
Yield-curve analysis becomes more useful when it documents failure modes openly. A low long-term yield can arise from central-bank bond purchases, safe-asset demand, demographic saving patterns, or a declining term premium. These factors may push the curve toward inversion without delivering the same message as a rapid rise in expected policy rates.
Likewise, a curve can steepen for opposing reasons. Short rates may fall because traders expect easing and slower activity, producing a bull steepening. Alternatively, long rates may rise because inflation uncertainty or term premiums increase, creating a bear steepening. Calling both events “economic recovery” would ignore the underlying mechanism.
| Analytical mistake | Why it fails | Better check |
|---|---|---|
| Assuming every inversion guarantees a recession | A historically useful predictor still allows false alarms | Review labor, credit, demand, and policy data |
| Using an intraday dip as a historic signal | Classic studies often use monthly averages | Check duration and averaging frequency |
| Ignoring the term premium | Long yields reflect risk compensation as well as rate expectations | Compare models and note uncertainty |
| Calling steepening universally bullish | Short-rate cuts and long-rate selloffs mean different things | Identify which maturity moved first |
| Mixing nominal and real rates | Inflation compensation changes the interpretation | Use compatible nominal or real curves |
| Equating sovereign and corporate curves | Corporate yields include changing credit spreads | Separate Treasury slope from credit risk |
There is also a cross-country problem. A 2019 Federal Reserve Bank of St. Louis analysis found that the historical predictive relationship varied among major economies. It reported more false positives in the United Kingdom and Canada than in some other countries and a weaker relationship in Italy. Thus, a U.S.-based forecasting rule should not be copied mechanically to another government bond market.
How the Yield Curve Affects Bonds, Banks, and Credit Markets
For a bond investor, changes in yields translate into changes in market prices. When yields rise, the market price of a conventional fixed-rate bond generally falls; when yields decline, that price usually increases. The size of the reaction depends on duration, coupon, and convexity. Investors can examine the mechanics of bond yields before interpreting a steepening curve as a portfolio opportunity.
Bank margins require more than a one-line explanation
Banks often fund a portion of their assets with deposits and other relatively short-term liabilities while holding loans or securities that mature later. Consequently, a compressed yield spread can complicate the economics of maturity transformation. Yet the actual effect on a bank depends on deposit pricing, hedges, floating-rate loans, liquidity buffers, credit losses, and competition.
Funding pressure can become a separate risk channel
Short-rate increases can raise refinancing costs for businesses, households, and leveraged investors. If funding becomes scarce, an otherwise orderly move in yields may coincide with poor trading conditions. Thus, analysts should distinguish the slope of the Treasury curve from market liquidity, including bid–ask spreads and market depth.
A Practical Five-Step Framework for Reading the Curve
Instead of reacting to an inversion headline, use a repeatable process that separates observation from interpretation. The following framework also makes historical comparisons easier to reproduce.
- Choose the curve and quote convention. Specify whether you use U.S. Treasury nominal par yields, zero-coupon yields, inflation-protected securities, or another market. Record the source.
- Calculate at least two spreads. Check 10y–3m and 10y–2y, then examine the five-year and thirty-year points for humps or unusual changes.
- Measure persistence. Compare today’s reading with weekly or monthly averages rather than treating one intraday tick as a completed economic signal.
- Separate expectations from premiums. Ask whether short-rate forecasts moved, term premiums changed, or both. Use multiple estimates when possible.
- Cross-check the economy. Compare employment, spending, inflation, bank lending, and credit conditions. Record which data support or contradict the curve’s message.
Imagine a persistent 10y–3m inversion alongside rising unemployment, weaker new orders, and tighter lending standards. Several independent signals would then point toward higher downside risk. In contrast, if employment remains resilient, lending continues, and the inversion results largely from a falling term premium, the interpretation becomes less straightforward.
The last step requires understanding policy responses. A central bank may reduce short-term rates as inflation eases, while government fiscal measures can affect demand on a different timetable. Those channels are explored in our comparison of fiscal and monetary policy. In both cases, a changed policy outlook may reshape the curve before official economic data visibly turn.
Case Study: A Long-Bond Selloff Can Look Like a Growth Signal
A revealing example comes from the Federal Reserve Board’s analysis of the second half of 2023. Researchers documented a sharp rise in the ten-year Treasury yield from below 4% to above 5%, followed by a decline to roughly 3.9%. Instead of treating the swing as pure news about economic growth, they compared multiple models of expected short rates and term premiums.
For an analyst, the practical lesson is to test two competing explanations: did the curve steepen because near-term rate expectations changed, or because investors demanded greater compensation for long-term uncertainty? Neither explanation alone proves a forecast, but the distinction helps identify risks that a simple spread chart cannot show.
Yield Curve FAQ
What is a normal yield curve?
A normal yield curve slopes upward because longer-term government securities yield more than shorter-term securities. The shape can reflect expectations for future short rates and term premiums. However, an upward slope alone does not guarantee that economic growth will continue.
What does an inverted yield curve mean?
An inverted yield curve means at least one shorter-maturity security yields more than a longer-maturity security of comparable credit quality. Investors often associate persistent Treasury inversions with elevated recession risk, but an inversion cannot establish the timing or certainty of a downturn.
Is the 10-year/2-year spread better than the 10-year/3-month spread?
Neither spread is universally best for every task. The 10y–2y spread is a familiar market benchmark, while influential New York and San Francisco Fed studies favor 10y–3m for historical U.S. recession forecasting. Analysts should also evaluate forward spreads and the complete curve.
How many basis points equal one percentage point?
One percentage point equals 100 basis points. A movement from 4.25% to 3.90% is a change of −0.35 percentage point, or −35 basis points. Basis points make small differences across maturities easier to compare.
Can the yield curve steepen during a slowdown?
Yes. If markets anticipate rate cuts, short-term yields can fall faster than long-term yields and produce bull steepening even as growth expectations deteriorate. Rising long-term term premiums can also steepen the curve for unrelated reasons.
Does the yield curve tell an investor which bond to buy?
No. The Treasury curve helps an investor compare maturity compensation, but bond selection also depends on duration risk, cash needs, credit exposure, taxes, inflation, and liquidity. A steep curve does not automatically make the longest maturity the best choice.
Key Takeaways
The yield curve is most valuable when analysts read it as a set of prices containing several economic signals, not as a single recession switch. The shape reflects expected future short-term rates, risk compensation, and trading conditions. Meanwhile, different maturity spreads can disagree, model estimates can differ, and persistent inversions carry more historical weight than brief intraday episodes.
For financial-market decisions, the disciplined approach is to identify the maturity pair, calculate the spread consistently, determine which yields moved, and confirm the message against independent indicators. That approach preserves the curve’s genuine forecasting value without pretending it offers certainty about future growth, bond returns, or central-bank decisions.